Considering the Risks and Benefits of Proton Pump Inhibitor Use

Friday, March 6, 2020

Paul Anderson, MD

Proton pump inhibitors (PPIs; e.g. Prilosec/omeprazole, Prevacid/lansoprazole, and others) are among the most commonly used medications, with prevalence estimates ranging between about five percent and fifteen percent of the adult population. Recently, there have been increasing concerns about the safety of long-term PPI use, leading many patients to abandon their PPIs in favor of H2 receptor antagonists (H2RAs; e.g. Pepcid/famotidine, Zantac/ranitidine, Tagamet/cimetidine) or to stop acid suppression therapy altogether. While the tenet of “first, do no harm” remains paramount in the practice of medicine, it is important to use caution when interpreting reports of risks associated with PPI use.

“It is important to use caution when interpreting reports of risks associated with PPI use.”

Two recent reviews (Therapeutic Advances in Drug Safety. 2018; 10:2042098618809927 and Mayo Clinic Proceedings, Volume 93, Issue 2, 240 – 246) have looked critically at the reported risks of PPI use, the magnitude of the risk, and the quality of the evidence. Reported risks include C. diff infection (CDI), dementia, chronic kidney disease, osteoporosis/risk of fracture, interaction with anti-platelet therapy, community acquired pneumonia, small intestinal bacterial overgrowth (SIBO), and micronutrient deficiencies (magnesium and vitamin B12).

Relative risks/odds ratios for these concerns are mild to moderate, ranging from about 1.2 to 1.7. Relative risk for SIBO has been reported to be higher, 2.26, but the clinical significance of this finding is unclear. For dementia, CDI, and fracture risk, the data are inconsistent  and, in some cases, conflicting. For chronic kidney disease, the quality of evidence is very low, and an observed increase in use of nonsteroidal anti-inflammatory drugs (NSAIDs) among PPI users may confound this association.
There does appear to be a causal relationship between PPI use and micronutrient deficiencies (magnesium and B12) and between PPI use and SIBO, though the significance of the latter is unclear. Vitamin B12 and magnesium deficiency risk is mild (RR of 1.65 and 1.44, respectively) and can be managed with occasional lab and clinical monitoring.
Regarding concomitant use of clopidogrel and PPI, there is no interaction between pantoprazole and clopidogrel, and the risk of other PPIs in patients on clopidogrel is not well defined. When in doubt in these patients, pantoprazole can be safely used.
There does not appear to be a meaningful risk of pneumonia among PPI users.
It is probably reasonable to monitor creatinine, complete blood count (CBC), and magnesium levels every one to two years, and B12 levels every five years in long-term PPI users. For bone density, bone mineral density (BMD) screening as well as calcium and vitamin D intake should follow national guidelines.
In summary, the risks that have been reported to be associated with long term PPI use are mild to moderate in magnitude (with relative risk/odds ratios mostly less than two) and the quality of evidence is low or very low. Moreover, these are generally uncommon adverse events, so the absolute risk remains small. This small risk should be balanced against the potential benefit of PPI use, which is easily the most effective nonsurgical treatment for symptomatic gastroesophageal reflux disease (GERD), has revolutionized the management of peptic ulcer disease, and
may reduce risk of cancer among patients with Barrett’s esophagus.
In patients who are not at high risk of serious complications of acid-related disease, H2RA use with famotidine or cimetidine can be substituted for PPI use if effective. In patients whose symptoms are not controlled on H2RAs, or who have risk for serious complications of acid-related disease (such as Barrett’s esophagus, recurrent or large peptic ulcers, severe esophageal peptic strictures, among others), long-term PPI use is often justified.

“Wholesale cessation of PPI use because of reported risks is unnecessary and likely to the detriment of many patients.”

PPIs should be used for appropriate indications at the lowest effective dose and for the shortest necessary period, but wholesale cessation of PPI use because of reported risks is unnecessary and likely to the detriment of many patients.